2007 Sep 11 - Tue

Recent Paper on Profitability of Technical Stock Trading

There is a recent, very readable paper from Stephan Schulmeister called The Profitability of Technical Stock Trading has Moved from Daily to Intraday Data. His abstract goes like this:

This paper investigates how technical trading systems exploit the momentum and reversal effects in the S&P 500 spot and futures market. The former is exploited by trend-following models, while the latter by contrarian models. In total, the performance of 2580 widely used models is analyzed. When based on daily data, the profitability of technical stock trading has steadily declined since 1960 and has become unprofitable over the 1990s. However, when based on 30-minutes-data the same models produce an average gross return of 8.8% per year between 1983 and 2000. These results do not change substantially when trading is simulated over six subperiods. Those 25 models which performed best over the most recent subperiod produce a significantly higher gross return over the subsequent subperiod than all models. Over the out-of-sample-period 2001-2006 the 2580 models perform much worse than between 1983 and 2000. This result could be due to stock markets becoming more efficient or to stock price trends shifting from 30-minutes-prices to prices of higher frequencies.

One of the interesting comments he makes is that contrarian strategies appear to be more profitable than do trending strategies.

In the article, the author offers up some possible reasons why technical trading is harder (but I should temper that remark and say that successful trading is more profitable with 'higher frequency' data--5 minute bars over 30 minute bars or daily data):

The decline in the profitability of technical trading based on daily data could be explained in two different ways. The "adaptive market hypothesis. (Lo, 2004; Neely-Weller-Ulrich, 2006) holds that asset markets have become gradually more efficient, partly because learning to exploit profit opportunities wipes them out, partly because information technologies steadily improve market efficiency (Ohlson, 2004). The second explanation holds that technical traders have been increasingly using intraday data instead of daily data. This development could have caused intraday price movements to become more persistent and, hence, exploitable by technical models. At the same time price changes on the basis of daily data might have become more erratic. This would then cause technical trading to become less profitable based on daily prices (but not on intraday prices).

Another interesting quote I came across regarding how everyone's trades get jumbled together, and what trader's think about it:

... traders have to form expectations about expectations of all other traders (Keynes. "beauty contest. problem).

[/Trading/TechnicalAnalysis] permanent link

2007 Feb 13 - Tue

Books on Financial Time Series Analysis

There is a course being presented on Financial Time Series Analysis by J. Michael Steele. There is a reading list titled An Eclectic Selection of Books Pertaining to Financial Time Series. I reprint it here just in case it goes away:

General References:

Chris Chatfield, The Analysis of Time Series: An Introduction (6th Edition), Chapman and Hall, New York, 2004.

This is perhaps the most widely required texts for time series courses at the level of our course. It does not focus specifically on financial series, but it provides one will a good general basis. It strikes a sensible balance between theory and practice.

N. H. Chan, Time Series: Applications to Finance, John Wiley and Sons, New York, 2004.

A straightforward text that develops the theory of time series a the level of our course. It is less encyclopedic than Zivot and Wang, and this makes it easier to read. This text is useful even though it does not fully engage the struggle required by an honest attempt to understand real-world financial time series.

James D. Hamilton, Time Series Analysis, Princeton University Press, Princeton New Jersey, 1994.

For many, the "big green book" is their main resource. Weighing in at just under 800 pages, it is arguably the most complete treatment of the theory of time series as it is currently applied in economics and finance. It is more mathematical than our course, but for students who expect to make time series a serious part of their professional tool kit, it is worth the investment.

Terence C. Mills, The Econometrics of Financial Time Series (second edition), Cambridge University Press, Cambridge UK, 1999.

This book is close to the level of our course, and it provides good supplementary reading. Chapter 5, Modelling Return Distributions is particularly relevant. Whereas Zivot and Whang devote their energy to reporting on models that are off current interest, Mills takes a more editorial point of view. This is also one of our aims.

C.W.J. Granger (editor), Modelling Economic Series: Readings in Econometric Methodology, Clarendon Press, Oxford, 1990.

This is a collection of essays by leading econometrician's. The book now shows signs of age, but some bits are timeless, such as Leamer's "Let's Take the Con out of Econometrics." If I had picked the subtitle, I might have chosen "Modelling is not (or should not be) for Sissies."

State Space Models:

J. Durbin and S. J. Koopman, Time Series Analysis by State Space Models, Oxford University Press, 2000.

This is book is at the level of our class and it provides as smooth an introduction to state space models as you are likely to find. The basic theory is developed without going overboard.

A. C. Harvey, Forecasting, Structural Time Models and the Kalman filter, Cambridge University Press, 1989.

This text is also at the level of our course, and it is also well worth your time. When I first looked at it I thought it was "too hard" for our class, but now I don't see what I thought was the problem.

M. West and J. Harrison, Bayesian Forecasting and Dynamic Models (2nd Ed.), Springer-Verlag, 1999.

This book is often referenced, perhaps more often than it is read. Its 680 pages make it a book that many need to reference but few need to digest. Once you have some experience with state space models, it becomes a useful resource which (ironically!) turns out to be less encyclopedic than one might hope.

Works with an Attitude:

David F. Hendry, Econometrics: Alchemy or Science (New Edition), Oxford University Press, Oxford, 2000.

This bravely titled collection of essays is well-worth dipping into, though I doubt that few readers will plow through all of the individual works. Certainly one of the attractive features of the book is its willingness to tackle some hard issues head-on. De minimus, it gives us a list of the problems that you will face.

Authors of academic papers often relegate their acknowledgment of the shortcomings of their work to their closing paragraphs, and, just as often, they suggest that the present defects will be remedied at a later date. The authors and the readers quietly conspire in their knowledge that no remedy is unlikely to be forthcoming.

Robert E. Rubin and Jacob Weisberg, In an Uncertain World: Tough Choices from Wall Street to Washington, Random House, New York, 2003.

Rubin's premise is that to think wisely about the world, one must think probabilistically. He does not suggest that explicit models must be used at every turn, but he does argue that leaders are nuts unless they explicitly consider multiple circumstances that have widely differing likelihood of coming to pass. The work is autobiographical, and it comes from a certain political perspective. Nevertheless, Rubin is about as nonpartisan as a person can be who has had access to the top levels of financial decision making. This is a nontechnical book, but reading it will enrich almost anyone's understanding of the potential and the limitation of probabilistic models.

Andrei Shleifer, Inefficient Markets: An Introduction to Behavioral Finance, Oxford University Press, Oxford, 2000.

This brief, efficient survey puts on the table all of the most important examples of situations where the Efficient Market Hypothesis is known to break. It sets forth many of the basic arguments both for and against the EMH in all its many flavors.